Origins of Aloe vera

[One Windowsill]

The full paper is available at http://www.biomedcentral.com/1471-2148/15/29" onclick="window.open(this.href);return false; and is a fascinating look at the possible historical spread of Aloe vera as a medicine and the dispersion of the Aloe genus over longer times.

New Scientist has an article https://www.newscientist.com/article/mg ... superstar/" onclick="window.open(this.href);return false;

They suggest that over-collection drove Aloe vera to extinction in its original and possibly restricted location. This is known to have happened to other famous medicinal herbs during the Roman Empire, such as silphion. Another fascinating observation is that leaf succulence is highly linked to the likelihood of medicinal use of a particular species of Aloe, whether the gel is used or not (see Additional File 5). In the case of Aloe vera the portability and shelf life of the leaf must have contributed to its choice as a useful medicinal herb for the trade routes of the Arabian peninsula.

I am confused as to why the authors say there are seven close relatives of Aloe vera (which they don't name directly) found in the Arabian peninsula when the trees seem to show either four or ten depending how far you take it back. Perhaps saxicola can enlighten me? The four most obvious are A. acutissima, bakeri, purpurea and yemenica. Are the other three A. benishangulana, mawii and schelpii?

A. acutissima and A. bakeri appear to be exclusively Madagascan rather than the Arabian suggested by the paper, according to both Tropicos and Succulentguide. A. purpurea appears to be from Réunion/Mauritius not Arabia. A. yemenica does appear to be from the Yemen on the Arabian Peninsula.

Just to complete my confusion A. benishangulana is from Ethiopia , mawii Malawi and schelpii, Ethiopia. The other five in the "Arabian" aloes do seem to be all from the Yemen (and some more widespread in Arabia).

[Saxicola]

I didn't have time to read it in depth (I'll try to later), but I skimmed through with your question in mind. Keep in mind I'm in no way an expert on Aloes, so I can't talk about how the taxonomy matches up with the phylogeny. Also, I didn't want to get bogged down in the materials and methods, so I'm trusting that they had properly identified samples and did their analyses in a proper way. Finally, I am trusting that you are correct in the distributions of the species you mentioned.

Based on what I read, if I were a reviewer I'd take issue with how confident they are in their hypothesis. I'd say that they are presenting what is probably the most likely hypothesis of where A. vera originated, but if I were giving it odds of being correct I'd say 60-70% likely. I'd want to be at least 90% certain before I'd present my findings with the certainty they do.

If A. bakeri, A. acutissima, & A. purpurea are from far away islands then that explains how the clade of A. vera + 10 other species becomes 7 arabian relatives. Where my criticism as a reviewer would come in is that they don't even so much as mention the three island species (please correct me if I missed it). More importantly, those species make up three of the five species in the primary A. vera clade itself. In other words, in that clade, we only have one species, A. yemenica, that we are certain is native to Arabia, three not native to it, and one of indeterminate origin.

Lets assume the ancestor to the large clade which includes the "Arabian clade" and the sister clade (A. dorotheae to A. trichosantha on the cladogram) came from continental Africa. That A. dorotheae to A. trichosantha clade seems to be predominantly continental African (again correct me if I'm wrong). So we can say they stayed around home (with home being a huge continent!) while the other clade (the one we care about) migrated.

Where did the ancestor of this migrating clade originate? If we say Arabia like the authors do and we assume their phylogeny at the species level is correct then we have to say that Aloe purpurea found its way to Mauritius and Reunion, then seperately the ancestor of A. bakeri & A. acutissima made its way to Madagascar. Considering those three islands are close to each other and very far from Arabia, it is quite impressive that two different migration events occured to put them there without any evidence of anyone from the clade having moved back to Africa, despite it being much closer. The other interpretation is that the ancestor of this clade dispersed from somewhere on the African continent to those islands, diversified there and one species migrated to the Arabian peninsula giving us A. yemenica. In this scenario there is no compelling reason based on phylogeny alone that A. vera comes from Arabia. It could have come from Madagascar or even Africa itself. What I'm really saying here is that both scenarios are about equally plausible and each require explaining some significant dispersal events.

I wonder why A. rubroviolacea isn't in this phylogeny.

Something else they don't seem to fully consider is that A. vera could be a man made species that never existed in the wild like corn. Maybe it was a form of A. yemenica that was selected for big leaves and high levels of medicinal properties. Being highly selected and separated from its wild ancestors it could have changed significantly enough for us to consider it a distinct species. Or it could have been some natural or artificial hybrid. Considering that A. vera is mostly sterile a hybrid origin makes sense. Because of the way DNA sequencing and phylogenetic analysis work you won't necessarily be able to tell a hybrid from a regular species, especially an ancient hybrid.

"Just to complete my confusion A. benishangulana is from Ethiopia , mawii Malawi and schelpii, Ethiopia." Those species come out sister to the entire "arabian" clade PLUS the A. dorotheae-A. trichosantha clade. So them being native to Africa doesn't really affect things one way or the other, besides supporting my hypothesis that the ancestor of that big clade was from Africa itself. When reading a phylogeny don't look at the proximity of the names, look at how the nodes are connected. The further left the connecting node the more distant the relation between the clades.

[One Windowsill]

I may be right to be confused, then.

Did you miss the black bar in Figure 3 marking the "aloes of the Arabian Peninsula"? The three Ethiopian/Malawi species are apparently included. Aloe sabaea is not included, strangely, though the name leapt out as Arabian, it is Yemeni and Saudi.

[Saxicola]

Yes I saw that bar but I did make the mistake of not paying attention that A. benishangulana, A. mawii, & A. schelpii were included. Assuming they are wrong with that bar, which is what you are suggesting, then it doesn't change anything I said. It just makes them wrong. Remember that I said I don't know a lot about the genus so I was relying on you when you said three species are from Madagascar & Mauritius and three more are from East Africa.

The more I look at this the more I wonder if any of those people are actually Aloe taxonomists, or if they were non-taxonomists, maybe working with a taxonomist but not an Aloe expert, to study A. vera origins as part of a bigger project.

I guess the question I should ask you is are you certain that those 6 species we've been discussing as not native, really are not native to Arabia? If the answer is yes then I can only see this as sloppy work that wasn't caught by any authors, reviewers, or editors.

Also, it seems we now have at least two Arabian Aloes not included in the study. A. sabaea and A. rubroviolacea. That would be understandable for some species only known in habitat, but both of them are commercially available.

[One Windowsill]

I am no expert but those are the areas given by either Tropicos, Succulentguide or both. Can you suggest a definitive source?

A. sabaea and at least one other endemic Arabian aloe (A. praetermissa) were included in the study but not identified as being in the "aloes of the Arabian Peninsula", that is what I meant by "not included".

This chap has 28 of the 42 Arabian peninsula endemic aloe species, so this study was only done with broad brushstrokes. https://www.flickr.com/photos/32378995@ ... 034245860/" onclick="window.open(this.href);return false;

[One Windowsill]

It just occurred to me to try the Red List:

A. benishangulana only known from the type locality near Assosa in Benishangul-Gumuz, Ethiopia. http://www.iucnredlist.org/details/201325/0" onclick="window.open(this.href);return false;

A. schelpei is found in central Ethiopia (Shewa region), extent of occurrence is below 600 km². http://www.iucnredlist.org/details/201343/0" onclick="window.open(this.href);return false;

A. mawii not yet assessed but the Kew Catalogue of Life gives Distribution as MLW, MOZ, TAN.
http://www.catalogueoflife.org/col/deta ... ec36d3e75a" onclick="window.open(this.href);return false;

A. acutissima and A. bakeri not yet assessed but the Catalogue of Life gives Distribution as MDG.
http://www.catalogueoflife.org/col/deta ... 709c0b4c97" onclick="window.open(this.href);return false;
http://www.catalogueoflife.org/col/deta ... 3f9a5db1d8" onclick="window.open(this.href);return false;

Likewise with A. purpurea except Distribution as MAU. http://www.catalogueoflife.org/col/deta ... 088d78b583" onclick="window.open(this.href);return false;

[Saxicola]

That seems good enough to me for distribution.

This really has me wondering if something that sloppy really made it in the paper. So I'll try to read it in detail this weekend to see if the authors talk about it somewhere that I missed. I suspect not though as I did a word search for the three island species and didn't get any hits in the paper.

I'm even more confused now that I looked up the lead author. She is a taxonomist and specializes in Aloe. I thought this would be some Ethnobotanist who didn't know or care about the rest of Aloe. I don't know how she could make the distribution mistakes. I may email her after I read the paper to get clarification.